Why the devs not making the windows builds, i don't get it ?
Do you want a windows build? I've ported a version that needs testing.
Topic: SILENTARMY v5: Zcash miner, 115 sol/s on R9 Nano, 70 sol/s on GTX 1070 - page 41. (Read 209314 times)
November 14, 2016, 03:21:53 PM
Do you want a windows build? I've ported a version that needs testing.
November 14, 2016, 02:08:30 PM
add devfee we can turn off and on ,if it's off it doesn't slow us down i'll keep it on a few days at a time. I'll put it use. I hope others would to.
November 14, 2016, 12:42:13 PM
Add cpu mining support for DevFee with 1 core mining on your T addresses.
Include option from 0 to 100% DevFee. I am aware that binary distributions are preferred due to speed and developers get more donations due to existance of devfee but mining 24/7 with 1 core to dev address or 1 hour at full speed every 24h will make huge difference in development speed.
P.S. We need solutions fast. Everyone with cpu/gpu power should donate ASAP for faster development. Other closed source projects already support 150+ S/s on highend AMD cards, but instability is the only frontier that stops everyone using silentarmy to switch there...As if nobody cares about Claymore's and others devfee but only speed - please donate by mining on silentarmy devs addresses. I will ask every active developer to put their address on the github page so every and each one of us donates for you!
Include option from 0 to 100% DevFee. I am aware that binary distributions are preferred due to speed and developers get more donations due to existance of devfee but mining 24/7 with 1 core to dev address or 1 hour at full speed every 24h will make huge difference in development speed.
P.S. We need solutions fast. Everyone with cpu/gpu power should donate ASAP for faster development. Other closed source projects already support 150+ S/s on highend AMD cards, but instability is the only frontier that stops everyone using silentarmy to switch there...As if nobody cares about Claymore's and others devfee but only speed - please donate by mining on silentarmy devs addresses. I will ask every active developer to put their address on the github page so every and each one of us donates for you!
November 14, 2016, 11:26:08 AM
Thanks genoil :-D
Edit: I could build the CUDA_SA now..let see if it works :-D
But it doesn't work :-D registers get spammed :-D takes 5min to unload the application
Edit: I could build the CUDA_SA now..let see if it works :-D
But it doesn't work :-D registers get spammed :-D takes 5min to unload the application
November 14, 2016, 11:11:12 AM
Hello Krnlx,
can you please post the complete kernel.cu somehow your snippets aren't as complete as I thought on first spot.Getting this error:
Error 3 error : no operator "^" matches these operands X:\Mining\sources\nheqminer-cuda-silentarmy\cuda_silentarmy\kernel.cu 496 1 cuda_silentarmy
uint4 loada = *(__global uint4 *)((__global char *)a + 4);
uint4 loadb = *(__global uint4 *)((__global char *)b + 4);
uint4 stor = loada ^ loadb;
Or wasn't it supposed for your cuda port?
can you please post the complete kernel.cu somehow your snippets aren't as complete as I thought on first spot.Getting this error:
Error 3 error : no operator "^" matches these operands X:\Mining\sources\nheqminer-cuda-silentarmy\cuda_silentarmy\kernel.cu 496 1 cuda_silentarmy
uint4 loada = *(__global uint4 *)((__global char *)a + 4);
uint4 loadb = *(__global uint4 *)((__global char *)b + 4);
uint4 stor = loada ^ loadb;
Or wasn't it supposed for your cuda port?
It's for opencl. Cuda have not native 128bit xor (don't know about amd, and future cards). For cuda you can test
uint4 stor;
stor.x = loada.x ^ loadb.x;
stor.y = loada.y ^ loadb.y;
Code:
static __device__ __forceinline__ uint4 operator^ (uint4 a, uint4 b) { return make_uint4(a.x ^ b.x, a.y ^ b.y, a.z ^ b.z, a.w ^ b.w); }
November 14, 2016, 11:02:55 AM
Hello Krnlx,
can you please post the complete kernel.cu somehow your snippets aren't as complete as I thought on first spot.Getting this error:
Error 3 error : no operator "^" matches these operands X:\Mining\sources\nheqminer-cuda-silentarmy\cuda_silentarmy\kernel.cu 496 1 cuda_silentarmy
uint4 loada = *(__global uint4 *)((__global char *)a + 4);
uint4 loadb = *(__global uint4 *)((__global char *)b + 4);
uint4 stor = loada ^ loadb;
Or wasn't it supposed for your cuda port?
can you please post the complete kernel.cu somehow your snippets aren't as complete as I thought on first spot.Getting this error:
Error 3 error : no operator "^" matches these operands X:\Mining\sources\nheqminer-cuda-silentarmy\cuda_silentarmy\kernel.cu 496 1 cuda_silentarmy
uint4 loada = *(__global uint4 *)((__global char *)a + 4);
uint4 loadb = *(__global uint4 *)((__global char *)b + 4);
uint4 stor = loada ^ loadb;
Or wasn't it supposed for your cuda port?
It's for opencl. Cuda have not native 128bit xor (don't know about amd, and future cards). For cuda you can test
uint4 stor;
stor.x = loada.x ^ loadb.x;
stor.y = loada.y ^ loadb.y;
November 14, 2016, 10:35:15 AM
Yeah, I think it's totally worth it to put up a serious open source competition against Sir. Claymore 
We shall be back.
I'll try get Claymore's v5/optiminer speed in 2-3 days.We shall be back.
p.s. he have a more motivation because his model (miner with fee) works perfectly - opposite to my (opensource miner and own pool), miners don't want support coinsforall.io
I had a rough time with your pool so I gave it up some time ago, but have no problem letting the miner mine for a dev fee, perhaps you can sync up with mrb and figure something out..
I like the way optiminer mines the dev fee, eg it continuously mines instead of dropping all the hash to mine for dev. That also gives the user a accurate picture of how much hash they are getting.
People don't like being stuck or forced to use a single pool, thats just science.
November 14, 2016, 10:22:50 AM
I'll try get Claymore's v5/optiminer speed in 2-3 days.
p.s. he have a more motivation because his model (miner with fee) works perfectly - opposite to my (opensource miner and own pool), miners don't want support coinsforall.io
p.s. he have a more motivation because his model (miner with fee) works perfectly - opposite to my (opensource miner and own pool), miners don't want support coinsforall.io
I would have mined on coinsforall.io from day one but your pool seems blocked from chinese IPs? chinese were complaining that you blocked them from your pool from day 1.. Is it you or your host? everyone in china was complaining that it was not fair for them,... and that the ZCash start was not fair due to your pool not allowing direct chiense connections.
it was a pain in the butt to connect through VPN in linux, my VPn keeps dropping when the screen turns off so i would have to always leave the monitor on.. and since I was just using one monitor for 3 rigs... was impossible
did/do you know about the issue with china IPs not connecting to your pool? is this done on purpose? if not can you ask for a change of IP from your host?
I liked your miner and was also very upset i was not able to efficiently use it on your pool... really upsetting for me
at least I left you a small tip as I was using it for a few hours until i gave up due to the IP address situation..... so you got my remaining balance from day 1 - which today isn't work jack but still,,., i lots hours of mining on your pool on genesis day
you may want to consider changing hosts - but maybe damage is already done and there are too many pools now
and anyways... how about adding a ZClassic Pool?
November 14, 2016, 10:18:24 AM
Yeah, I think it's totally worth it to put up a serious open source competition against Sir. Claymore
We shall be back.
I'll try get Claymore's v5/optiminer speed in 2-3 days.We shall be back.
p.s. he have a more motivation because his model (miner with fee) works perfectly - opposite to my (opensource miner and own pool), miners don't want support coinsforall.io
I had a rough time with your pool so I gave it up some time ago, but have no problem letting the miner mine for a dev fee, perhaps you can sync up with mrb and figure something out..
November 14, 2016, 09:55:12 AM
Hello Krnlx,
can you please post the complete kernel.cu somehow your snippets aren't as complete as I thought on first spot.Getting this error:
Error 3 error : no operator "^" matches these operands X:\Mining\sources\nheqminer-cuda-silentarmy\cuda_silentarmy\kernel.cu 496 1 cuda_silentarmy
uint4 loada = *(__global uint4 *)((__global char *)a + 4);
uint4 loadb = *(__global uint4 *)((__global char *)b + 4);
uint4 stor = loada ^ loadb;
Or wasn't it supposed for your cuda port?
can you please post the complete kernel.cu somehow your snippets aren't as complete as I thought on first spot.Getting this error:
Error 3 error : no operator "^" matches these operands X:\Mining\sources\nheqminer-cuda-silentarmy\cuda_silentarmy\kernel.cu 496 1 cuda_silentarmy
uint4 loada = *(__global uint4 *)((__global char *)a + 4);
uint4 loadb = *(__global uint4 *)((__global char *)b + 4);
uint4 stor = loada ^ loadb;
Or wasn't it supposed for your cuda port?
November 14, 2016, 09:50:13 AM
Yeah, I think it's totally worth it to put up a serious open source competition against Sir. Claymore
We shall be back.
I'll try get Claymore's v5/optiminer speed in 2-3 days.We shall be back.
p.s. he have a more motivation because his model (miner with fee) works perfectly - opposite to my (opensource miner and own pool), miners don't want support coinsforall.io
What do you think about FOSS miner with dev fee. Small fee 1-2% split among devs (address list in sources with weights). If someone builds one on its own he has chance to opt out. No enforcing, no saying that people opting out are bad.
There probably won't be much profit in that but some motivation is better than none, as someone that does OSS from begging and currently works for FOSS company I know that well.
Also I am planning to rewrite computation pipeline to always have some job queued up thus reducing kernel downtime. The overhead is small right now but when the limitations of GPUs are truly met it might make a difference.
November 14, 2016, 09:12:57 AM
Yeah, I think it's totally worth it to put up a serious open source competition against Sir. Claymore
We shall be back.
I'll try get Claymore's v5/optiminer speed in 2-3 days.We shall be back.
p.s. he have a more motivation because his model (miner with fee) works perfectly - opposite to my (opensource miner and own pool), miners don't want support coinsforall.io
November 14, 2016, 09:08:38 AM
Build instructions for Windows:
1. Install AMD APP SDK for Windows from here http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/
2. Get cygwin from http://cygwin.com
3. Install cygwin with the packages:
- Devel/git
- Devel/make
- Devel/gcc-core
- Python/python3
4. Start cygwin64 terminal
5. Clone repo: git clone https://github.com/mbevand/silentarmy.git
6. cd silentarmy
7. Build: make OPENCL_HEADERS="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/include/" LIBOPENCL="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/lib/x86_64"
8. Copy cygwin1.dll: cp /cygdrive/c/cygwin64/bin/cygwin1.dll .
9. Get your build path by typing: cygpath . -w -a
10. Create start.bat with:
c:\cygwin64\bin\python3.4m silentarmy --instances=1 -c stratum+tcp://address -u x.x -p x --use 1,2
pause
It doesn't work, 3 dev show:"0 sol/s". 1. Install AMD APP SDK for Windows from here http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/
2. Get cygwin from http://cygwin.com
3. Install cygwin with the packages:
- Devel/git
- Devel/make
- Devel/gcc-core
- Python/python3
4. Start cygwin64 terminal
5. Clone repo: git clone https://github.com/mbevand/silentarmy.git
6. cd silentarmy
7. Build: make OPENCL_HEADERS="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/include/" LIBOPENCL="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/lib/x86_64"
8. Copy cygwin1.dll: cp /cygdrive/c/cygwin64/bin/cygwin1.dll .
9. Get your build path by typing: cygpath . -w -a
10. Create start.bat with:
c:\cygwin64\bin\python3.4m silentarmy --instances=1 -c stratum+tcp://address -u x.x -p x --use 1,2
pause
Do I installs AMD APP SDK 64bit and 32 bit? I only install AMD APP SDK 64bit
November 14, 2016, 09:03:42 AM
Does the Cygwin port also have issues wit multiple instances like my MSVC port?
Yes, using cygwin results the same performance issue when running multiple instances
November 14, 2016, 09:01:27 AM
Why the devs not making the windows builds, i don't get it ?
November 14, 2016, 08:54:40 AM
Well krnlx has a cuda port of it so I assume its in the kernel.cu since ther eis no input.cl in his cuda implementation. Or I'm blind since i found these functions in the kernel .cu too,
ah.. got it. as I have no cuda (as I'm using amd opencl drivers) I've just adjusted input.cl in my case. Sorry for confusing.
November 14, 2016, 08:48:36 AM
these changes are made in the kernel.cu? Am I right
input.cl
Well krnlx has a cuda port of it so I assume its in the kernel.cu since ther eis no input.cl in his cuda implementation. Or I'm blind since i found these functions in the kernel .cu too,
November 14, 2016, 08:46:22 AM
these changes are made in the kernel.cu? Am I right
input.cl
November 14, 2016, 08:40:44 AM
November 14, 2016, 08:34:02 AM
Build instructions for Windows:
1. Install AMD APP SDK for Windows from here http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/
2. Get cygwin from http://cygwin.com
3. Install cygwin with the packages:
- Devel/git
- Devel/make
- Devel/gcc-core
- Python/python3
4. Start cygwin64 terminal
5. Clone repo: git clone https://github.com/mbevand/silentarmy.git
6. cd silentarmy
7. Build: make OPENCL_HEADERS="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/include/" LIBOPENCL="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/lib/x86_64"
8. Copy cygwin1.dll: cp /cygdrive/c/cygwin64/bin/cygwin1.dll .
9. Get your build path by typing: cygpath . -w -a
10. Create start.bat with:
c:\cygwin64\bin\python3.4m silentarmy --instances=1 -c stratum+tcp://address -u x.x -p x --use 1,2
pause
1. Install AMD APP SDK for Windows from here http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/
2. Get cygwin from http://cygwin.com
3. Install cygwin with the packages:
- Devel/git
- Devel/make
- Devel/gcc-core
- Python/python3
4. Start cygwin64 terminal
5. Clone repo: git clone https://github.com/mbevand/silentarmy.git
6. cd silentarmy
7. Build: make OPENCL_HEADERS="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/include/" LIBOPENCL="/cygdrive/c/Program\ Files\ \(x86\)/AMD\ APP\ SDK/3.0/lib/x86_64"
8. Copy cygwin1.dll: cp /cygdrive/c/cygwin64/bin/cygwin1.dll .
9. Get your build path by typing: cygpath . -w -a
10. Create start.bat with:
c:\cygwin64\bin\python3.4m silentarmy --instances=1 -c stratum+tcp://address -u x.x -p x --use 1,2
pause
Thank you very much for this.
Jump to: